Go Down

### Topic: need help to find a low flow rate sensor with at least 10Hz samping rate (Read 2185 times)previous topic - next topic

#### tony_arduino

Hi All,
I am trying to build a system to quantify the boundary condition of a fluid system with low flow rate : about 200ml/min. And most important is that I need the system to give me a 'real time' measurement(say, the sampling rate is at least 10 Hz to give me a useful result) of the information of flowing liquid, say, the speed or flow rate or pressure. I checked flow rate sensor online but the response time is too low, which is about 1s. And the price is above \$200 for the sensor.
Do you have any idea?
Thanks.
Tony

#### Grumpy_Mike

What fluid, are you measuring all of it like a gas pipe or just a bit like a river?

200ml/min, wow, that is very little!
Just 3-4ml/sec.
Is that the kind of volume you count in drips, or is there some discernible flow there?
Designing & building electrical circuits for over 25 years.  Screw Shield for Mega/Due/Uno,  Bobuino with ATMega1284P, & other '328P & '1284P creations & offerings at  my website.

#### robtillaart

What is the diameter of the hose / pipe /tube (right word) ?

IIRC it is possible to measure flow by changes in electric field due to flow (sorry no link, it will probably depend on the liquid)

Quote
Just 3-4ml/sec.

Quote
the sampling rate is at least 10 Hz to give me a useful result

that means samples of 0.3 ml??

According to wikipedia: one drop of water is ~ 50 microliter.

So were talking about counting 3 ml/sec = 60 drops per second  Think that is too fast to make some drop counting device, 60 drops is a continuous jet allready I think.

If it are still drops maybe a laserbundle on a photocell?

What is the range of flow that must be measured?
minimum speed:
maximum speed:

What is an acceptable error?

Rob Tillaart

Nederlandse sectie - http://arduino.cc/forum/index.php/board,77.0.html -
(Please do not PM for private consultancy)

#### retrolefty

Quote
IIRC it is possible to measure flow by changes in electric field due to flow (sorry no link, it will probably depend on the liquid)

I know with water (and some other but not all fluids) there are magnetic flow meters that require no direct contact with the process being measured, but I'm pretty sure they won't work well at this desired low flow rate and small piping size. Ultrasonic flow meters are also of a non-wetted kind, but again not at such small pipe and low flows. Some kind of positive displacement flow meter is probably what you will end up with, that generates a pulse for each given mass quanity that flows through it. Count the pulses for a given specific time and you get flow rate. Think paddle wheel with a magnetic pick-up to count the flow rate. I'm sure you won't find anything real affordable, unless you happen to find some surplus lab type meter on e-bay. If you can run your process in a 'batch' mode, there is always letting the fluid flow into a vessel until the weight of the vessel reaches a given value and then release the fluid to the process. Strain gauges (load cells) can be sized for very small values and are pretty accurate.

Lefty

#### robtillaart

Good idea lefty,
That means you could also weigh the source of the liquid  if it is some bucket or so

Another thing is, why does the liquid flow?
- is it a pump (maybe we can measure something at the pump)
- is it gravity and a faucet? (maybe we can measure something at the faucet)
- can we add something to the liquid that passes by ? (paddle wheel as lefty said)

Why you want to measure this? Can you tell us more about the greater system, its goal?

Rob Tillaart

Nederlandse sectie - http://arduino.cc/forum/index.php/board,77.0.html -
(Please do not PM for private consultancy)

#### mjbmikeb

I agree, weight is a good option.
Start with a \$10 load cell from a kitchen scale and see if the signal to noise at 10 readings a second is good enough for your needs.
Run everything from batteries and shield using a Faraday cage.
If your experiment is repeatable then sum multiple runs to give a better signal to noise ratio.

#### tony_arduino

Hi,
Sorry for the short discreption. For this project, we are mimicking the blood flowing in a tube, so the velocity is low, we use a pump to push the blood mimicking fluid flow into a tube(4mm). So the flow rate is as low as our blood.
Blood flow rate through each carotid artery = 350 ml/min (Kandel et al., Principles of Neural Science, New York: McGraw Hill, 2000)
Blood flow rate through basilar artery = 100-200 ml/min (Kandel et al., 2000)
For the velocity, should between ?1/6)m/s to (1/2)m/s.
I will measure the fluid property during its steady state, so I pump it from 1 pot to another pot.
Thanks.

Tony

#### robtillaart

Quote
we use a pump to push

Measure if the pump is on/off would be a start..

Quote
I will measure the fluid property during its steady state, so I pump it from 1 pot to another pot.

So two scales one under each pot (you need two to detect leakage;)
Rob Tillaart

Nederlandse sectie - http://arduino.cc/forum/index.php/board,77.0.html -
(Please do not PM for private consultancy)

#### retrolefty

#9
##### Jul 20, 2011, 06:18 pmLast Edit: Jul 20, 2011, 06:20 pm by retrolefty Reason: 1
Quote
I will measure the fluid property during its steady state, so I pump it from 1 pot to another pot.

For this to work accuractly pump must be of the positive displacement type and you much insure all tubing between pots and pump suction remain liquid packed at all times. Gravity can be your friend, be sure pump suction has the lowest physical placement of the setup.

Lefty

#### tony_arduino

#10
##### Jul 20, 2011, 06:33 pm
Hi,
I had asked for the turbine flow rate sensor or the positive displacement sensor in Cole Parmer but their response time is about 1s which is too low. I think the second way you posted will be more suitable to my case though I haven't use the strain gauge before. So I need some time to do research on it.
Thanks again.

Go Up